Classification of Chocolate Consumption Using Support Vector Machine Algorithm

Abstract

Chocolate, derived from the processing of cocoa beans (Theobroma cacao), is a widely consumed product with potential health risks when consumed excessively. This study investigates the classification of chocolate consumption behaviors using the Support Vector Machine (SVM) algorithm and evaluates its classification performance. A benchmark dataset on chocolate consumption was employed, partitioned into nine folds for training and testing purposes. To mitigate issues related to data imbalance, the Synthetic Minority Over-sampling Technique (SMOTE) was applied. The experimental findings indicate that SVM, enhanced by SMOTE, demonstrates a reliable capacity for classifying chocolate consumption categories. Performance evaluation across multiple experiments revealed variations in Accuracy, Precision, Recall, and F1-Score, with overall accuracies ranging from 50% to 60%, suggesting moderate but consistent classification performance.